REPORT:

STATES AND LOCALITIES ARE LOSING MONEY ON AIRBNB’S TAX DEALS

Airbnb’s “voluntary” tax agreements are short-changing state and local
governments by millions of dollars in tax revenue
By Dan Bucks
Former Director, Montana Revenue Department/Executive Director, Multi-State Tax Commission
June 2019

Airbnb’s “Voluntary” Tax Payments are Uncertain, Illusory and Unreliable

Airbnb regularly takes credit for having produced certain amounts of revenue for various states and cities—often the numbers
run into millions of dollars. It does not, however, produce a comprehensive annual report for amounts submitted to all states
and localities with which it has agreements. Regardless of the amounts, what we know is that at best the payments Airbnb is
taking credit for are uncertain, illusory and unreliable. Worse yet, Airbnb’s secrecy business model impairs the administration
of a variety of taxes and thus causes significant, unacceptable losses of revenue at all levels of government.

State And Localities Are Losing Millions Of Dollars In Tax Revenue From Airbnb’s
“Voluntary Agreements”
The “Median State” table below presents a comprehensive view of the impact the Airbnb tax agency agreements and Airbnb’s
1099 reporting practices have on major federal, state and local tax revenues. Because it is based on the midpoint of actual
data from every state and DC, the table is more representative of all states than any single state.

The table’s results are for a model state with median levels of: population, urban/rural shares of that population, and tax rates.
The taxes include state sales tax on lodging, local lodging taxes, federal income taxes from this state, state income tax, and
local property taxes.

This table can be used to estimate revenue losses for states and the nation overall. For example, after adjusting for different
tax systems among the states, the “median state” revenue loss of $32.45 million in 2018 translates into a total national
loss of federal, state, and local revenues of $1.64 billion in 2018 and $3.48 billion over the 2013-2018 period.

Median State Example:

Revenue Impact of Airbnb Agreements and 1099 Reporting (millions of dollars)
Category 2013 2014 2015 2016 2017 2018 Cumulative
State Sales Tax Loss (0.44) (0.88) (1.61) (4.69) (8.90) (12.46) (28.98)
Airbnb Payments 0 0 0 4.83 9.18 12.85 26.85
Net Sales Tax Impact (0.44) (0.88) (1.61) 0.14 0.28 0.39 (2.12)

Local Lodging Tax Loss (0.73) (1.46) (2.60) (7.80) (14.82) (20.75) (48.17)
Airbnb Payments 0 0 (2.68) 8.04 15.28 21.39 47.39
Net Lodging Tax Impact (0.73) (1.46) 0.08 0.24 0.46 0.64 (0.78)

Federal Income Tax Loss 0 0 (1.06) (3.17) (6.03) (8.44) (18.70)

State Income Tax Loss 0 0 (0.30) (0.90) (1.71) (2.40) (5.31)
Total Income Tax Loss 0 0 (1.36) (4.08) (7.74) (10.84) (24.02)

Property Tax Loss (0.38) (0.97) (2.08) (7.39) (15.09) (22.64) (48.56)

Total Revenue Loss (1.55) (3.32) (4.97) (11.08) (22.20) (32.45) (75.47)
Notes: Median state population of 4.6 million, and median urban population of 74% of total state population; Airbnb earnings data for 2016 from March 2017 CBRE report
scaled to median state size.; Growth rates used to estimate other years estimated from hotelappraisers.com study, April 2018; State Airbnb agreement assumed to start in
2016; Local Airbnb agreements assumed to start in 2015 and are in place in all localities; Median sales tax is 6%, median lodging tax is 13.5%, middle-income federal tax
rate is 17.6% and state tax rate is 5%, median property tax rate differentials are 0.575% (2013), 0.585% (2014), 0.594% (2015), 0.625% (2016), 0.659% (2017 & 2018);
Federal income tax loss is for income earned by lodging operators within a state only; Income tax losses attributable to reduced compliance behavior due to secrecy of
income beginning in 2015; Property tax losses caused by property appraisers unable to identify short-term rental properties. Totals may not add due to rounding.
These nationwide estimates are a measure of the revenue damage done by Airbnb’s agreements as aided by tax
agencies signing them—damage that may be expanding as agencies extend Airbnb agreements to other lodging
marketplaces. They are solid estimates based on mid-point data carefully analyzed. Even if cut arbitrarily and unjustifiably
by a third, the losses still top $1 billion for 2018 and $2.3 billion over five years. If Airbnb criticizes these numbers, the reply
is simple: End the secrecy. Release the data to tax agencies—the earnings, the records behind the payments, the identity
and location of the lodging facilities, and more so that actual losses can be calculated. Launching criticisms from behind a
wall of secrecy that hides relevant evidence is not credible or reasonable.

THE FACTS:
Airbnb has fueled its rapid growth since 2014 by offering commercial-style lodging operations using entire homes
and apartments converted from residential use to transient lodging—operations that often violate local housing and
zoning policies. To prevent any of these illegal rentals from being closed, Airbnb works to keep the identity and location of
its lodging operations secret from public officials. That secrecy has also encouraged operators to avoid collecting sales and
lodging taxes and paying income taxes and has shield them from commercial property tax assessments.

Airbnb hardened the shield of secrecy for its lodging operators by offering to make payments, ostensibly for lodging
and sales taxes, in return for highly questionable, special treatment by tax agencies—including being allowed to keep
their lodging operators secret from all public authorities. Unlike other taxpayers, Airbnb is also not accountable for the
payments made under the agreements. Airbnb has secured such agreements in over 35 states and several hundred localities.

The terms of the Airbnb agreements not only create uncertainty about the payments made, but they are also
illusory because while they create a false impression of revenues gained, they also facilitate revenues lost. The
secrecy provisions of the agreements contribute to suppressing property taxes in a majority of states and, in conjunction with
Airbnb’s 1099 reporting practices, also decreases federal, state and local income tax compliance and revenues. While
Airbnb’s business model creates these revenue losses everywhere, the states or localities that have signed the Airbnb
agreements guarantee that they will experience the full force of these losses.

Only a sliver of Airbnb’s activities represents new or increased tourist visits (two to four percent at most) to a state
or city.1 Most Airbnb revenue is simply a diversion of lodging stays from traditional lodging sources, which generally collect
and pay lodging taxes through a more transparent and documented tax collection process. Thus, whatever money Airbnb
collects and sends to a tax agency under an agreement does not result in any significant net gain in lodging tax revenue.

An overwhelming share of Airbnb’s tax payments only make up for the lost revenue caused by Airbnb recruiting
short-term rental operators who typically do not comply with sales and lodging taxes and whose identities Airbnb keeps
hidden from public agencies at the outset of the booking process. In the best-case scenario, states or cities that enter collection
agreements may receive payments from Airbnb that cover sales and lodging tax revenue losses caused by Airbnb’s business
model in the first place.

The choice of the word “may” here is deliberate, because whether the amount of these payments equal what is actually due
is uncertain under the agreements. However, even in the best-case regarding sales and lodging taxes, total governmental
revenue—federal, state and local—would decline because Airbnb’s secrecy shield and 1099 income reporting practices
decrease revenues below what the law requires from Airbnb lodging operators for income taxes—federal, state, and local—
and for property taxes where commercial property rates are higher than residential rates.

Airbnb is not accountable for its payments because its agreements typically specify that tax agencies cannot audit
the full books and records of Airbnb. These agreements allow Airbnb to only provide anonymous data on the transactions—
data that hides the name and location of the lodging operator. These unprecedented and highly irregular restrictions on tax
audits make it virtually impossible to verify the accuracy of Airbnb’s payments.

1 See Josh Bivens, “The economic costs and benefits of Airbnb,” Economic Policy Institute, for a review of the literature on this subject. His analysis
concludes that 96% to 98% of the spending related to Airbnb lodging would have occurred even if Airbnb did not exist. Travelers would have stayed at
traditional lodging sources instead. His analysis refuted an Airbnb commissioned study that implausibly assumed, without evidence, that travelers using
Airbnb took all their trips only because Airbnb existed.
Manatee County, Florida, Tax Collector Ken Burton described these
anonymous data audits well when he compared them to: “Manatee County Tax Collector
Ken Burton compared arguments that
. . . an income tax software manufacturer telling the IRS it Airbnb should collect and remit the
will pay taxes owed by its customers without revealing their local bed tax on behalf of all of its
identities. He noted that the IRS would not permit such an clients in Manatee to the idea of an
arrangement and his office cannot, either.2 income tax software manufacturer
These audit restrictions effectively transfer the power to telling the Internal Revenue Service
determine what taxes will be paid from the tax agency to Airbnb. that it will pay taxes owed by its
That, in turn, creates substantial uncertainty as to whether Airbnb customers without revealing their
payments equal the right amount of tax. Airbnb can issue all the identities. He noted that the IRS would
press releases it wishes about the payments it has sent to states and not permit such an arrangement and
localities. However, those press releases cannot guarantee that what his office cannot, either.”
it paid equals what the law requires.

On average, states tax commercial property—such as entire Herald-Tribune, 8/21/2017

homes rented out through Airbnb— are 64 percent more
heavily than residential property.3 These practices vary widely among the states. However, over thirty states have
effective rates for commercial property that range from 10 percent to 300 percent more than residential property. If tax
assessors do not know that a residence has been converted into a short-term rental, they will continue to apply the lower
residential rates to the property. In states or localities where Airbnb agreements exist, the secrecy provisions make it
virtually impossible for property tax assessors to identify those short-term rentals that should be reclassified from residential
to commercial property and taxed accordingly.

The result is a nationwide property tax revenue impact estimated at $763 million in 2018 and $1.636 billion in the
five-year period from 2013 through 2018.4 This large effect is not surprising given that property taxes are the largest
revenue source for local governments. However, the impact is not uniform around the nation because of the wide variation
among the states in property tax practices. 5 Overall, the property tax loss caused by the Airbnb agreements is nearly half of
the total estimated national revenue loss for all taxes and easily overwhelms the small net revenues from Airbnb’s collection
of sales and lodging taxes.

The damage to property taxes is likely greater than the estimated five-year $1.636 billion revenue loss. As Airbnb
lodging facilities gain market share from traditional lodging sources, the assessed value of those traditional lodging facilities
either declines or slows its growth. So, the property tax losses from the traditional lodging sector are also not being
recouped by appropriate commercial taxation of full-time Airbnb rentals. The amount of property tax revenue foregone or lost
from downward pressure on traditional lodging values is difficult to measure. However, this additional factor underscores the
conservative nature of the property tax estimate.

The importance of this property tax issue has often been overlooked in discussions of the public impact of the
increase in short-term lodging rentals. Both individual state studies and a national review are certainly overdue.
Governments that have signed Airbnb agreements and that also tax commercial property at higher effective rates than
residential property should place a high priority on determining how these agreements affect their property tax revenues.
Unfortunately, that is huge challenge because the secrecy provisions of the agreements make it exceedingly difficult to
identify properties now taxed as residences when they should be commercial property.

2 Quoted in Dale White, “Manatee seeks individual compliance with resort taxes,” Saratoga Herald-Tribune, August 21, 2017.
3 Lincoln Institute of Land Policy and Minnesota Center for Fiscal Excellence, “50-State Property Tax Comparison Study: For Taxes Paid in 2018,” April
2018, p 34.
4 Depending on state and local budgeting laws, the property tax revenue impact can arise, in the short-term, as either a revenue loss or a tax increase

shifted to other property taxpayers. The report assumes that the impact translates into a loss for simplicity purposes and because, overtime, legal or
political limits on raising property tax rates converts these impacts into chronic losses. In the either case, the dollar amount of the impact is the same
whether its effect is taken out of government budgets or the budgets of other taxpayers
5 The median state example uses, as it should, the average differential in tax rates between residential and commercial properties. That results in the

property tax impact being the largest in the median state case for all types of taxes. However, when converting the median state example to a
nationwide estimate, accounting for states with either no differential or a partial differential reduces the national property tax impact to a level slightly
below the total federal and state income tax impact.
The potential income tax losses arise, in the first instance, from Airbnb’s practice of supplying 1099 income reports
to only those lodging operators earning more than $20,000 a year. Airbnb is allowed to do so under special rules that
apply to the 1099-K documents that it provides. However, some electronic marketplaces—most notably Lyft—voluntarily
provide their participants with 1099 reports for all those earning only $600 or more. The CBRE study found that in 2016 the
average annual earnings for all Airbnb lodging operators was $13,674. 6 Thus, a significant majority of Airbnb operators likely
do not receive a 1099 report. The IRS has determined that when taxpayers do not receive 1099 or comparable reports, the
rates of voluntary tax compliance fall by 56 percent, from 93 percent compliant to only 37 percent.7 That impact is across the
board for all income taxes—federal, state, and local—except for those taxpayers in Massachusetts and Vermont. 8

The 56 percent reduction in voluntary income tax compliance for Airbnb rental earnings yields a nationwide
combined income tax revenue loss estimate of $735 million in 2018—$599 million in federal income tax losses and
$136 million in state losses. Over the five-year, 2013-2018 period, the total income tax losses are estimated at
$1.628 billion—$1.327 billion federal and $301 million state.

The Airbnb agreements likely decrease the level of income tax compliance below 37 percent. Because of the growing
awareness that these agreements keep secret from state tax agencies not simply their earnings but also the very existence
of their lodging business, lodging operators are even more likely to fail to report their lodging rental income on their tax
returns. Beyond simply the absence of 1099 reports, the impact of this active shielding of businesses from the tax agencies
has not been measured. However, it is reasonable to assume that the more hidden an economic activity is, the less likely it
will be voluntarily reported on income tax returns.

Airbnb could readily remedy the problem of lost income tax revenues by simply reinstating the approach it used
before tax year 2015 and that is still followed by Lyft and other online sellers of voluntarily providing 1099 income
reports for all its lodging operators that earn more than $600 in rental income. 9 The alternative course is for states to adopt
laws mandating a $600 reporting threshold as Massachusetts and Vermont have already done.

A third problem with the perceived revenue gains from Airbnb agreements, is that the payments received under
those agreements are unreliable in the sense that, unlike true tax payments, the governments cannot depend on the
continuing level or even existence of the payments. This conclusion arises from both the restrictions on audits cited
above and the cancellation clauses contained within the agreements. 10 These provisions convert the Airbnb payments from
being considered taxes into a discretionary or optional contract payment.

The agreements give Airbnb the right to cancel on short notice—thirty to ninety days—and cease the payments for
any reason even though its economic activity in the jurisdiction continues. Also, the audit restrictions in the agreement
grant Airbnb the ability to reduce future payments regardless of what was actually collected from lodging guests or the level
of lodging receipts attained. The definition of “taxes” does not allow taxpayers discretion to cancel or artificially reduce
payments even though economic activity continues forward at the same or higher pace. If the economic activity continues,
taxes need to be filed and paid in accordance with the law and not at the whim of the payor. In short, these provisions make
the payments optional, instead of mandatory. Thus, the Airbnb payments likely do not qualify as taxes under governmental
accounting rules.

There are real fiscal consequences that arise from these payments not qualifying as taxes. If the payments are
subject to Airbnb’s discretion and cancellation, they should be excluded from tax revenues for budgeting
purposes. They should also be excluded from taxes in the governments’ official accounting statements. That, in turn, can

6 CBRE, Supra at note 4, Table 1, p. 5.

7 Internal Revenue Service, “Federal Tax Compliance Research: Tax Gap Estimates for Tax Years 2008-2010,” at https://www.irs.gov/pub/irs-
soi/p1415.pdf. See Chart 1 at 12.
8 Massachusetts and Vermont have enacted relatively new state laws that require 1099 reports to be provided to everyone earning more than $500 a

year, thus eliminating the impact of the federal 1099-K rules.

9 Airbnb stopped reporting to operators earning more than $600 when it began actively pursuing its secrecy agreements with tax agencies. Unknown if

Airbnb decided sending tax agencies required copies of 1099s with lodging operator names and earnings would undercut secrecy of VCAs.
10 The first report noted that the audit restrictions were one factor in disqualifying the Airbnb agreements from being treated as tax settlement

agreements. However, that report did not note the impact of the cancellation clause on these issues. Both features of the Airbnb agreements disqualify
(a) the agreements from being considered tax agreements, and (b) the payments from being considered taxes.
 affect the credit rating of these governments. Finally, calculations of bonding capacity also cannot consider these payments
to be taxes. The notion that payments cancellable on a few weeks or months’ notice might be pledged as backing for
repayment of long-term bonds is questionable if not foolish. For all these purposes the Airbnb payments should be
accounted for as some type of optional contract payments that are less reliable than tax revenues.

Independent auditors should review the accounting, budgeting and bonding treatment given to Airbnb payments
against governmental accounting standards. Those auditors should advise state and local governments of any corrective
actions those governments need to make. More generally, elected officials should reevaluate the wisdom of accepting at
face value breezy statements from a fast-growing, public relations-savvy company that says, “We want to pay taxes.”11 If
those statements sound too good to be true, that just might be the case.

Conclusion

Airbnb’s “voluntary” tax agreements, when all relevant tax sources are considered, lose more public revenue than they
gain from Airbnb’s payments, and the more Airbnb grows, the more federal, state and local revenue is lost.

11 Chris Lehane of Airbnb quote in Martineau, Supra at note 17.